“Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Chris O’Hara, co-founder and chief revenue officer at Bionic Advertising Systems.
There has been a lot of talk about the pervasive amount of click fraud and bot traffic happening in digital. Marketers are reportedly spending anywhere from 30% to 70% of their digital budgets on fake impressions and clicks, and an entire cottage industry is cropping up to help marketers combat fraud and protect their digital marketing investments.
Some people claim that the price of fraud is already built into the programmatic RTB ecosystem. Marketers use programmatic RTB for direct marketing and measure sales using CPA metrics. If they pay $100 per verified acquisition, should they care whether it takes 10 million or 20 million impressions to produce a conversion? Some say that they don’t and take the view that they only pay for results, justified by backend conversion metrics that take media cost into consideration.
I hope this is not the case. Ignoring fraud with these justifications is what ultimately may kill the digital advertising business before we ever get to scale.
Let’s take the case of the big programmatic marketing platform that has been getting great conversions for their customers. Marketers look at the results of such platforms and think that the technology has managed to effectively separate the wheat from the chaff in popular ad exchanges and find the “sweet spot” of cookie targeting that converts. But dig a little deeper and you notice that many of the conversions are happening on webmail subdomains, such as mail.yahoo.com. In other words, the platform is getting last-view attribution from successful email marketing. This is a more subtle case of fraud – really more of a tax on digital ignorance for marketers. But again, the marketer sees this channel producing results that align with his CPA goals.
Did the conversions get attributed correctly? Maybe not, but those questions get overlooked when the blended CPA is on target.
Cookie bombing and other types of fraud are just as likely to limit digital advertising to performance budgets and keep real growth at bay.
If we are being honest with ourselves, we must admit that there doesn’t seem to be a ton of desire to solve these inherent problems in programmatic RTB. There are too many people making too much money to want to fix it. And it’s going to destroy programmatic RTB as we know it.
Who benefits from the current scenario?
Publishers: Most publishers benefit greatly from the programmatic RTB revenue stream. Big publishers “fill” their long-tail inventory with ads. Mid-sized publishers without large direct-sales teams depend greatly on network and programmatic fill for their revenue. Long-tail pubs are fully committed to their AdSense checks for survival. A lot of publishers’ comScore numbers are a lot bigger than they should be, thanks to cheap inventory of unknown provenance.
Ad tech: Every vendor in programmatic RTB benefits from inventory flowing through their pipes. Most charge on a percentage of spend, which means they might sacrifice 50% of their revenue if they had to stop charging for fraudulent impressions. New fraud-detection and measurement firms also profit, albeit in a virtuous way.
Agencies: What would today’s big agencies do without the ability to leverage programmatic RTB to arbitrage inventory, or charge a premium for “unpacking the ad tech space” for clients? The new programmatic landscape has been a boon to smart, nimble agencies that have built, bought or leveraged ad technology to pivot their dying media businesses. How eager are agencies to expose the fundamental flaws within the programmatic RTB ecosystem?
The Biggest Loser
The biggest loser in the entire room is the poor marketer, who ultimately pays the bills. Yet it’s easy to turn a blind eye because the numbers look good. But how long will big marketers confuse true marketing success with today’s flawed digital attribution metrics?
Marketers are starting to think about real measurement frameworks – net new customers – rather than CPA metrics. They are also keenly interested in using brand messages to interact with their customers across screens. And they won’t be using CPA to measure brand growth.
So why do marketers continue to leverage programmatic RTB despite the inherent risk of fraud and current limitations for brand advertising?
To paraphrase Clear Channel’s Bob Pittman at the recent IAB annual meeting: “Given a choice between quality and convenience, convenience always wins.”
The biggest question lately is whether or not we can make it as convenient – and cheap – to buy guaranteed media at scale. Seeing this opportunity, a lot of players in programmatic RTB are looking hard at the money being spent on guaranteed media – the “transactional RFP” channel – and trying to add new “programmatic direct” tools to their arsenals. RTB players know that brands are still uncomfortable executing campaigns in the wilds of the open exchange, and they know truly premium inventory won’t be available unless publishers have more granular control over pricing, availability and partner selection.
There are some exciting new technologies out there to combat fraud, and even some transformational new data-driven approaches to stopping fraud before it happens. A6 Corp’s Internet Topology, for example, successfully illuminates the dark corners of the bot-driven Internet so marketers can avoid fraud before it even begins by avoiding bidding on bot traffic.
The real key to winning the battle against fraud and closing the Meeker Gap for good is going to be whether or not marketers insist on standards for inventory quality and stop leveraging the faulty attribution metrics we have created for them. I recently spoke with a technology company that powers real-time biding for hundreds of demand-side platforms. They see up to 50% of bot traffic on digital display inventory, and up to a stunning 65% on online video inventory. They have also been developing powerful tools that help combat click and impression fraud online. Sadly, they were surprised to see very little adoption of the tools.
Is it surprising that technology that would cause demand-side platforms to lose 50% of their transactional revenue overnight is not popular?